AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Robust Learning with Progressive Data Expansion Against Spurious Correlation

Yihe Deng, Yu Yang, Baharan Mirzasoleiman, Quanquan Gu

NeurIPS 2023 · Architectures and efficiency › CNN design · approach: Invertible and wide-network analysis · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add progressive data expansion stage after warm-up (ablation)PDE · Waterbirdsworst-group accuracy: 86.0 -> 90.3Improved
add progressive data expansion stage after warm-up (ablation)PDE · CelebAworst-group accuracy: 87.8 -> 91.0Improved
add progressive data expansion stage after warm-up (ablation)PDE · CivilCommentsworst-group accuracy: 67.7 -> 71.5Improved
add all new data at once instead of gradually (ablation)PDE · synthetic dataworst-group accuracy: 93.01 -> 74.24Hurt
reset momentum after warm-up (ablation)PDE · synthetic dataworst-group accuracy: 93.01 -> 92.51Hurt
reset momentum after warm-up and add all data (ablation)PDE · synthetic dataworst-group accuracy: 93.01 -> 67.69Hurt
expansion size 5 or 50 instead of 10 (ablation)PDE · Waterbirdsworst-group accuracy: 90.3 -> 88.1Hurt
expansion learning rate 1e-2 instead of 1e-4 (ablation)PDE · Waterbirdsworst-group accuracy: 90.3 -> 85.4Hurt
expansion learning rate 1e-3 or 1e-5 instead of 1e-4 (ablation)PDE · Waterbirdsworst-group accuracy: 90.3 -> 89.4No clear effect
train GroupDRO with the larger learning rate used by PDEGroupDRO · Waterbirdsworst-group accuracy: 86.7 -> 77.3Hurt
PDE vs GroupDRO convergence speedPDE · Waterbirdsearly-stopping epoch: 92 -> 8.9Improved
PDE vs GroupDROPDE · Waterbirdsaverage accuracy: 93.2 -> 92.4No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy